Why not? Matt Yglesias objects to my post yesterday, holding it up as an example of why nobody likes economists. He says that my analysis ignores the fact that “big part of the point of prostitution prohibition laws is to express social disapproval of prostitutes and prostitution”:
Indeed, people seem generally quite unconcerned about whether prostitution is occurring someplace out of sight and out of mind. But they want to reserve the right to strongly disapprove of both the prostitution and especially the prostitutes. You can analogize a person who engaged in a form of sexual or commercial conduct of which you disapprove by referring to that person as a “whore.” It’s an insult. Its insult status reflects and upholds a social consensus that whores are bad people, not just that whoring is a kind of undesirable nuisance. Side-payments can’t address this issue.
The first thing I would point out is that he seems to agree with me with respect to the actual welfare analysis of the voluntary exchange of prostitution. This at the very least suggests that there should be a pareto improving legal framework where the only effect of the law is that any visible signs of prostitution are banned, including advertising, streetwalking, etc. This would demonstrate social disapproval while allowing exchange. It also suggests that given that there are anti-prostitution laws, the optimal enforcement, with the exception of the public nuisance element of streetwalkers (which I’ll talk about later), should be zero.
Is this what we observe now? Crackdowns on very secretive and high-priced prostitution rings do occur, and Rhode Island recently passed a law that moved the state from legalized behind-closed-door, no advertising, prostitution to full prohibition.
The other, arguably more important, issue raised here is that I was specifically addressing externalities of prostitution. Should people’s preferences over laws regulating some good or service be considered externalities of that good or service? I think the answer here is clearly no, since they are unrelated to the quantity of the service produced.
However, one might still argue that these preferences should be included in welfare analysis as a cost or benefit of the law. This is an interesting question.
One concern here is that voters have preferences over most laws, and as Bryan Caplan argued in The Myth of the Rational Voter, those preferences suffer from fundamental biases not related to the costs and benefits of those laws. He documents 4 biases:
1. Make-work bias
2. Anti-foreign bias
3. Pessimistic bias
4. Anti-market bias
If we are to accept that voters preferences over laws -as opposed to preferences over the direct outcomes of those laws- should be included in cost benefit analysis, then we should be prepared to consider laws which satisfy voters inherent anti-foreign bias as a legitimate benefit.
One could argue that public policy, and cost-benefit analysis therein, should not weigh the enjoyment some get from lowering the status of others. When weighing protectionist measures, for instance, should we consider the fact that they satisfy anti-foreign bias as a benefit?
Then again, one could make an argument that the costs and benefits of the expressive value of laws should be taken into consideration. After all, if laws lower the status of some group, and they have negative disutility from that which they’d be willing to pay to remove, then that is a real cost. When framing the problem as one of harming some group like this it seems more worthy of consideration. But both costs and benefits must be considered. This would mean weighing the cost of lower status of prostitutes against the benefits of those who wish to lower their status. Likewise one could weigh the benefits of satisfying anti-foreign bias against the costs to those of us who find that bias reprehensible.
Matt seems to think the reason people don’t like economists is because they miss these things or they ignore them. That’s a fair enough criticism (although again, in my case I was addressing the particular question of externalities). But he should consider Robin Hanson, whose willingness to wrap any cost or benefit into welfare analysis, no matter how egregious, is surely the purest form of economic thinking. No offense to Robin, but I don’t think people would like economists more if they all conducted economic analysis more like him.
More on streetwalkers later.
10 comments
Comments feed for this article
Friday ~ November 5th, 2010 at 10:02 am
jazzbumpa
In the three relevant posts, you talk a lot about mental abstractions. This is fine, in and of itself, but there ought to be more to it than just an intellectual exercise. Are you leading up to a policy recommendation?
FWIW, I don’t think Sis Y’s idea has much merit.
Sister Y argues that prostitution in fact has an externality. Some individuals find prostitution morally objectionable, and so they suffer mental costs when someone else hires a prostitute.
There are legitimate externalities to consider, which I won’t enumerate here. That mental cost, however, is not among them. This is someone’s damned, self-imposed problem; and they can just deal with it. I’m assuming here that the individual is a neighbor, co-worker, or some random guy who might not even know Mr Hirer. BTW, isn’t Renter more appropriate?)
It’s a bit different for Mr. Hirer’s wife, though. I’d say she has a vested and legitimate concern.
And that’s where this kind of analysis can get very sticky, indeed. Who actually has a legitimate concern?
Certainly not somebody whose delicate sensitivities are ruffled, or who wants to dictate morality to another adult.
Cheers!
JzB
Friday ~ November 5th, 2010 at 2:50 pm
Wonks Anonymous
That’s also an argument against Robert Frank’s view of the negative externality of affluent peers.
Monday ~ November 15th, 2010 at 1:07 pm
Sister Y
I was not referring to mental externalities, but to the potential for religious folks’ imaginary friends raining down fire on us for fornicating. I mean, can you PROVE their imaginary friends don’t exist?
Friday ~ November 5th, 2010 at 11:08 am
Econ Skeptic
The totally retarted part about the economic thinking of this sort is that it regards people’s preferences as a static and independent constant. Just plug as many of these measured constants into the equation and you’ll inform your thinking as to what the Pareto-optimal thing to do is. The reality, of course, is that preferences do move, including in response to laws and the incentives that laws (and enforcement of laws) provide. And this preference-based economic analysis has no way to account for that (behavioral models are somewhat better). It’s such an obvious shortcoming in discussion of things generally considered to be morally reprehensible or in some ways repugnant and which may also happen to be illegal at some point in time, that it’s totally shocking that you could have written two long posts about it without noticing a glaring hole like that in the analysis.
Next thing you know, we’ll be arguing that the abolution of slavery was not pareto optimal at the time and therefore should not have happened.
Friday ~ November 5th, 2010 at 2:52 pm
Wonks Anonymous
Pareto-optimality is too strict, most economists go for Kaldor-Hicks without investigating whether side-payments actually take place.
Saturday ~ November 6th, 2010 at 9:29 pm
Econ Skeptic
Not sure what the point of this reply is. The notion of pareto-optimality is clearly not cental to the point I was making.
And, to answer you directly, those economists that WOULD apply the Kaldor-Hicks criterion to measuring the efficiency of things like prohibition on prostitution, drugs, gambling or – yes – slavery are even more insane than those who try the analysis in the first place in the manner suggested by the author of the post. I mean, talk about ignoring reality for the sake of doing some useless static-equilibrium gymnastics.
Friday ~ November 5th, 2010 at 11:09 am
Overcoming Bias : Disapproving Via Bans
[...] One could … [argue] that the costs and benefits of the expressive value of laws should be taken into consideration. .. This would mean weighing the cost of lower status of prostitutes against the benefits of those who wish to lower their status. … Matt seems to think the reason people don’t like economists is because they miss these things or they ignore them. That’s a fair enough criticism. But he should consider Robin Hanson, whose willingness to wrap any cost or benefit into welfare analysis, no matter how egregious, is surely the purest form of economic thinking. No offense to Robin, but I don’t think people would like economists more if conducted economic analysis more like him. (more) [...]
Friday ~ November 5th, 2010 at 1:01 pm
kharris
If we are going to consider a system in which utility is attached to satisfying bias, then we need to comprehend every bias in that system. You have made reference only to blue-nose bias. I find that, even when I am not drinking or screwing, I get a lift from knowing that others are.
It is clearly not novel to suggest that internal externalities – externalities that have only to do with my prefernces, and do not otherwise affect my welfare – be taken into account. Politicians live on just such notions. Far better to gratify my prejudices than to provide me a better world. Taking bias into account, though, is not something we ought to make routine and stop thinking clearly about. Reasonably, (though impractically – which is always going to be the case when trying to tally up biases and the utility of gratifying them) we’d want to do the analysis with and without bias. If the two results point in the same direction, fine. If they don’t then we need to think twice about lowering the material welfare in order to reward bias.
Friday ~ November 5th, 2010 at 1:57 pm
Psychohistorian
The analysis seems completely wrong. I’ll go in order of descending wrongness.
First, there isn’t any practical way to apply this, because there’s no way to actually assess who would be prostitutes and determine their reservation prices. How would we actually negotiate so we would be giving money only to those who would otherwise be working as prostitutes? How would we enforce it? We already aren’t that good at catching people doing it, so how would we catch people who broke their contracts?
We don’t know the prostitute’s reservation price. If it’s $380, then it would only cost $20 to have him refrain from prostitution for one hour. This substantially alters the balance, and means that your cost is exaggerated. When you consider that the hourly wage of the typical street prostitute is closer to $30, the problem seems even smaller.
And on that note, the compensation scheme is absurd. You wouldn’t pay an hourly rate. When you get a non-compete agreement as part of your severance package, you don’t get paid for every hour you’re not going to work (getting another job that doesn’t violate it does not change your compensation, even though it would alter the benefit of competing). You get a lump sum. You wouldn’t pay people per-hour not to prostitute themselves, you’d give them a lump sum annual or monthly payment.
Indeed, if it weren’t for the enforcement problems, you could probably pay a hooker making $20,000 a year significantly less than she makes in order to get her to stop hooking. Again, the problem is there’s no way to spot which women would actually be hookers and which just want free money for abstaining from something they wouldn’t be doing anyways.
You also make a time-valuation mistake. I worked in a business with high hourly pay but with only about 20 hours of work a week available. It would not have made sense for me to value my leisure time at my hourly rate, because I wanted to be working a lot more but could not. Similarly, prostitutes cannot scale their work indefinitely. The demand for prostitutes at 9 am on a Wednesday is decidedly lower than at midnight on a Saturday. This is a relatively minor point in light of the other ones.
So, if you could specifically select the people who wanted to be prostitutes, you could probably pay most of them a lump sum that people would be willing to subsidize. But then you’d have to be able to enforce it. And you’d have to be able to prevent everyone who wasn’t a prostitute from taking payment for abstaining from something they didn’t want to do anyways. It’s kind of like eliminating laws against assault and allowing people to barter to not get punched in the face. Some people could extract a large amount of money for abstaining from activities they would otherwise never have considered.
This is not a problem markets could solve, though of course I disagree with the basic premise that we have a right to be bothered by the consenting acts of other adults.
Friday ~ November 5th, 2010 at 2:11 pm
Psychohistorian
Paying people not to do drugs would be a better example than not to be violent. Lots and lots of non-users would cash in.